A systems approach to routing global gridded runoff through local high-resolution stream networks for flood early warning systems

Abstract

Global or national scale flood early warning systems (FEWS) can benefit developing countries and ungauged regions that lack observational data, computational infrastructure, and/or the human capacity for streamflow modelling. Existing land surface models (LSM) typically generate forecasts using coarse resolution grid cells which, at least for streamflow, have little value when used for flood warning at local scales. We present the design and development of a new automated computational system, using existing, well-established open source software tools that quickly downscales (or maps) the runoff generated from such coarse grid-based LSMs onto high-resolution vector-based stream networks then routes the results using a vector-based river routing model. We conducted experiments using the ERA-Interim/Land reanalysis data – a 35-year retrospective gridded runoff data product from the European Center for Medium-range Weather Forecasts (ECMWF) – to assess our fast downscaling system. The accuracy of our approach is comparable to the Global Flood Awareness System (GloFAS) – a well-established gridded routing model using the same forcings – but our method provides streamflow predictions on significantly higher resolution stream networks. We found that the river network resolution has negligible effect on the simulated streamflow with our model routing. In other words, we can forecast streamflow for very small stream segments and potentially improve local flood awareness and response much more successfully than previously possible using readily available climate forcings from LSMs.